The Analysis of the Effectiveness of Fiery Semantic Segmentation Model
Name
Ilmar Möls
Abstract
This Bachelor's thesis focuses on analysing the performance of the Fiery model in detecting people from frontal camera images of a self-driving car using the Fiery model. The accuracy of the perception system of self-driving cars is critical, as it helps to ensure both road safety in traffic with such vehicles and the efficient operation of the vehicle's control system.
The thesis covers what the Fiery model is and what is the ability of this model to identify people in the case of different persons or conditions affecting the image. In addition, the paper provides an overview of the scientific background related to self-driving cars, including the safety issues associated with autonomous vehicles and the positioning techniques used. In conclusion, the Fiery model seems to have detection problems with pedestrians on the edges of images and pedestrians lying on the road.
The thesis covers what the Fiery model is and what is the ability of this model to identify people in the case of different persons or conditions affecting the image. In addition, the paper provides an overview of the scientific background related to self-driving cars, including the safety issues associated with autonomous vehicles and the positioning techniques used. In conclusion, the Fiery model seems to have detection problems with pedestrians on the edges of images and pedestrians lying on the road.
Graduation Thesis language
Estonian
Graduation Thesis type
Bachelor - Computer Science
Supervisor(s)
Markus Kängsepp
Defence year
2024